discrepancy between EV3IRsensor and NXT Ultrasonic

This is where you talk about the EV3 software itself, installation issues, and programming talk.

Moderators: roger, gloomyandy, skoehler

discrepancy between EV3IRsensor and NXT Ultrasonic

Postby steveiswicked » Tue Nov 12, 2013 9:23 pm

Hi, the output of the EV3IRSensor and the NXT Ultrasonic Sensor seem to deviate in a non linear way. I've tried calibrating them against a ruler and the Ultrasonic Sensor seems more trustworthy, could you give any advice on how to bring these values into line?

EV3IRSensor UltrasonicSensor Difference PilotDistance
54 60 -6
56 60 -4
55 62 -7
57 64 -7
59 67 -8
59 69 -10
59 69 -10
60 69 -9 -9.983284
60 70 -10
61 72 -11
62 74 -12
63 76 -13
64 77 -13
64 79 -15
65 79 -14
64 79 -15 -10.0112095
65 79 -14
65 80 -15
66 81 -15
66 82 -16
67 84 -17
68 86 -18
70 87 -17
68 88 -20
69 90 -21
69 90 -21
72 90 -18
70 90 -20 -10.0112095
69 91 -22
70 91 -21
70 92 -22
70 255 -185
72 255 -183
72 96 -24
72 98 -26
73 99 -26
75 100 -25
74 100 -26
73 100 -27
74 100 -26 -9.983284
73 100 -27
73 101 -28
73 102 -29
74 102 -28
76 104 -28
76 105 -29
76 106 -30
78 108 -30
76 109 -33
77 109 -32
77 109 -32
77 109 -32 -10.0112095
78 110 -32
78 110 -32
78 110 -32
78 111 -33
78 112 -34
78 114 -36
78 255 -9.983284
steveiswicked
New User
 
Posts: 19
Joined: Sun Feb 06, 2011 11:22 pm

Re: discrepancy between EV3IRsensor and NXT Ultrasonic

Postby Aswin » Tue Nov 12, 2013 10:42 pm

Hi, can you tell some more about the circumstances of the measurements. We're these taken from a moving robot or was the robot stationary while each measurement was taken? Also, what kind of object was used on the measurements? What is the meaning of the fourth column in the output? Were the measurements of both sensors taken at the same time?

To my experience the EV3 US sensor is quite precise. I think lego claims that the error is less than 3%. The errors I experienced were less and strongly correlated to the distance to target.

Aswin
My NXT blog: http://nxttime.wordpress.com/
Aswin
leJOS Team Member
 
Posts: 188
Joined: Tue Apr 26, 2011 9:18 pm
Location: Netherlands

Re: discrepancy between EV3IRsensor and NXT Ultrasonic

Postby gloomyandy » Tue Nov 12, 2013 11:59 pm

I don't think that Lego claim that the IR sensor output is linear. The following note is from the Lego on line documentation:
Code: Select all
The Proximity value does not correspond directly to a specific distance. The value will depend on the color and material of the object in front of the sensor, and other factors.


You might find the following site (that Aswin pointed out to me a little while ago) of interest:
http://www.ev-3.net/en/archives/848
User avatar
gloomyandy
leJOS Team Member
 
Posts: 3646
Joined: Fri Sep 28, 2007 2:06 pm
Location: UK

Re: discrepancy between EV3IRsensor and NXT Ultrasonic

Postby steveiswicked » Wed Nov 13, 2013 5:42 pm

Thanks Andy & Aswin, the description seems to tally with my observations. For the record EV3IR and NXT US sensor were front mounted above and below each other alongside a compass sensor, the robot runs a threaded client that continuously obtains sensor readings and sends them to a server app on the PC which plots them on a map. The robot was programmed to gently move backward away from a wall in 10cm steps at which point the readings would be reported while it was stationary (that's the fourth column, the traveldistance reported by the pilot). I'd hoped to combine the US and IR readings in a kalman filter to obtain a better overall reading, but it doesn't look like there's going to be any benefit.
steveiswicked
New User
 
Posts: 19
Joined: Sun Feb 06, 2011 11:22 pm


Return to EV3 Software

Who is online

Users browsing this forum: No registered users and 3 guests

more stuff